Michael Fremer at Rega.

[Tony L]

Most tests I've seen seem to indicate that Regas run indeed slightly fast out of the factory to compensate for the tendency to slow down as they age.

It's inaudible apparently.

My TD-124, like most Mk IIs that have been fully stripped down, cleaned and relubed with modern oils, starts a little fast and gradually slows down to speed after it has been running for an hour or so. On start up there is a very obvious yet smooth and even drift of the strobe, though even as someone with a pretty decent sense of pitch it is inaudible. It is speed variations that are hugely annoying (e.g. eccentrically pressed records, of which there are a disturbing amount), not absolute pitch. At worst we are talking a few cents here and there, e.g. fractions of a semi-tone. Useful pitch calculator here. As an example if you type in 100Hz and 101Hz, i.e. a 1% speed difference that equates to 17 cents, there are 100 cents in a semi-tone. Many guitars and basses will have more error than that in intonation.

 

[Robert]

Very good points there.

 

[drummerman]

My TD-124, like most Mk IIs that have been fully stripped down, cleaned and relubed with modern oils, starts a little fast and gradually slows down to speed after it has been running for an hour or so. On start up there is a very obvious yet smooth and even drift of the strobe, though even as someone with a pretty decent sense of pitch it is inaudible. It is speed variations that are hugely annoying (e.g. eccentrically pressed records, of which there are a disturbing amount), not absolute pitch. At worst we are talking a few cents here and there, e.g. fractions of a semi-tone. Useful pitch calculator here. As an example if you type in 100Hz and 101Hz, i.e. a 1% speed difference that equates to 17 cents, there are 100 cents in a semi-tone. Many guitars and basses will have more error than that in intonation.

No problem. I didn't mean a continuous very small increase in speed is audible

DC servo motors often have some drift. Can get bad with component ageing.

My 115 usually drifted for a few seconds but was otherwise stable.

 

[YNWOAN]

Not directly relevant to the Rega discussion (sorry), but I would just like to thank Paul R for his development of his 'real time' speed measurement protocol. For my coleagues and I Paul's work has been ground breaking and allowed us to study speed continuity and stability issues in a way hitherto impossible.

 

[yuckyamson]

That sounds awesome. Can you guys describe the method in a way that a luddite such as myself would understand? (I'm being genuine)

 

[Paul R]

That sounds awesome. Can you guys describe the method in a way that a luddite such as myself would understand? (I'm being genuine)

Basically you digitally capture a test tone from a turntable, then use a DSP FM demodulator on the recording. The output is effectively speed against time at a high resolution, so you can see the effect of the poles of a direct drive, the eccentricities of an idler, or motor pulley, the 50/100Hz pulsing of a synchronous motor, etc. Or the errors carved into each revolution of the test disc....

Rega are doing something similar with an optical pickup, which opens the possibility of capturing the turntable's performance while playing a real recording. They didn't go into the real stuff in Fremer's video, but the FFT of the speed variation was a clue.

I'm asking a chap I know who's an F1 sub-contractor where one might go for the encoder. Poking about the Internet hasn't offered much. I think the pickup side of the equation is relatively easy.

Paul

 

[yuckyamson]

Okay so being able to measure it to whatever accuracy is one thing....is the goal then to be able to better measure or actually help with implementing into the drive systems you're building/Power supplies?

 

[Tony L]

Or the errors carved into each revolution of the test disc....

This being a huge issue IME. Amazingly given the word 'test' in the title I've yet to see a properly concentric audiophile test record. HFS69, HFS75 and the Hi-Fi News ones are all eccentrically pressed to some degree. I guess any slight warps may bevan issue too given the resolution. Certainly enough to make these LPs useless for such a test unless you can somehow dial it out in software without losing any similar once per revolution wow in the drive system which may be present. The Rega system having a fixed position optical pickup would negate this entirely I guess assuming it is reading radial lines.

PS I tried using a chromatic guitar tuner with one of these records at one point, can't remember which, and it is interesting to see how even slight degrees of eccentricity can impact pitch (I can easily hear it too). As I mention above this is way, way more important than absolute pitch and certain highly collectable and expensive Nakamichi decks aside there is nothing that can be done about it. It would be interesting to try and do the maths at some point (well beyond me!), i.e. try and figure how many cents a mm of eccentricity equates to.

 

[Paul R]

Okay so being able to measure it to whatever accuracy is one thing....is the goal then to be able to better measure or actually help with implementing into the drive systems you're building/Power supplies?

It was/is about exploration of what's going on in a turntable. And possibly determining whether a modification has the effect you predicted.

This being a huge issue IME. Amazingly given the word 'test' in the title I've yet to see a properly concentric audiophile test record.

It should be possible to fix this by hand, the grooves will be concentric, just not to the existing centre hole. So drill it out, fill it up and make a new one....

I think it's possible to detect and correct off-centre pressings in a digital environment. So you rip the LP, pass it through the magic and it comes out centred, and pitch corrected if you wanted. I think there's been some work with analogue tape replay to pick up the bias signal on replay and use that to correct speed from moment to moment. This takes out the W&F of both the recorder and player. We could probably work some easy magic on a CD4 pressing, but that's a rather limited application.

Paul

 

[Paul R]

It would be interesting to try and do the maths at some point (well beyond me!), i.e. try and figure how many cents a mm of eccentricity equates to.

Groove speed is proportional to radius, so the impact of the constant offset is dependent on where on the disc you are playing.

So, for example, at a radius of 140mm with a note of nominally 440Hz and an centering error of 1mm it will wobulate between 436.86 and 443.14 every 1.8 seconds. I think that's about 12 hundredths of a semitone at that frequency.

(Happy to be corrected here.... Stupid mistakes are expected.)

Paul

 

[yuckyamson]

Cleveland Symphony Orchestra tunes to A-436, so....

Whatever.

I just hope the Naiad runs accurately.

 

[lsmanley]

I found the videos very interesting indeed.

I do suspect now that anyone wanting a s/h Aria is going to want to know if it's the newer version now or the better RP8/RP10 bearings though!

Well now I'm thinking exactly that, having invested in an Aria (brand new) about a year ago!!

 

[Jim Audiomisc]

Groove speed is proportional to radius, so the impact of the constant offset is dependent on where on the disc you are playing.

So, for example, at a radius of 140mm with a note of nominally 440Hz and an centering error of 1mm it will wobulate between 436.86 and 443.14 every 1.8 seconds. I think that's about 12 hundredths of a semitone at that frequency.

(Happy to be corrected here.... Stupid mistakes are expected.)

Paul

Not checked your figures. But my reaction to the topic is to wonder about the typical and actual levels of centering error on commercial LPs - and how often this swamps the wow from the actual turntable.

Variations at higher rates are a different issue, I assume. But even there I wonder what the performance is of lathes. I can't recall ever reading values for this. Add in possible variations due to drag when replayed and I wonder at what point trying to reduce wow may not matter given other factors?

 

[barryb]

My new RP6 runs fast - 33.7. FWIW, I hadn't read of this issue prior to purchase, though I did notice it immediately on first play, which prompted me to take the trouble to measure it. That said, it doesn't bother me particularly. To truly enjoy vinyl, one must be stoic!

 

[hifi-dog]

Mines at 33.7 also...

 

[barryb]

Some cursory reading around indicates 33.6/7 seems to be the norm. The addition of an Edwards Audio Blue Belt seemingly remedies this. I've bought one though haven't bothered installing it. The replay of vinyl is inseverable from its inherent benefits and weaknesses, and to these ears, this is fallibility I can bear; indeed, it's perhaps additive in some cases! Ultimately, coming from a generation essentially raised on digital, I've learned to make peace with audio gremlins to enjoy vinyl.

 

[Seanm]

I do believe the RP6/Exact also sounds fast even operating at spec: when I first heard mine I was sure the speed was way out, but it measures fine. Something to do with relative emphasis on attack over decay? Dunno. I like it though.